Medical image analysis (2008/2009)

Course partially running

Course code
Name of lecturer
Andrea Giachetti
Number of ECTS credits allocated
Academic sector
Language of instruction
3° Q dal Apr 20, 2009 al Jun 19, 2009.

Lesson timetable

3° Q
Day Time Type Place Note
Tuesday 9:30 AM - 11:30 AM lesson Lecture Hall F  
Friday 10:30 AM - 1:30 PM practice session Laboratory Alfa  
Friday 10:30 AM - 1:30 PM lesson Lecture Hall D  

Learning outcomes

The aim of this course is to introduce some image processing and visualization techniques not studied in the other courses and mainly applied to the medical field.
In detail, modalities to acquire digital diagnostic images will be analyzed as well as protocol to store and transmit the data.
Algorithms to process 2D and 3D images will be then described and practical applications will be performed using Matlab.


1. Digital images and diagnostic imaging.

Goal: a review of image processing and an overview on images in hospitals.
-Digital images and related processing.
-Diagnostic imaging modalities: CT, MRI, US, PET, ecc.
-DICOM: image communication and archive in medicine
-Overview of medical image applications: Computer Aided Diagnosis, surgical planning, simulation
- Volume data visualization, Surface and Volume rendering techniques

2. 3D data segmentation and visualization.

Goal: Describing the most used 3D-4D recosntruction and visualization used in the medical practice
-Thresholding, region growing, mathematical morphology
-Methods based on clustering in color space, Graph cuts, Watershed, MRFs
-"Snakes" and other 2D/3D deformable models
- Model based approaches
- Region/volume processing, feature extraction, distance functions, curve skeletons

3. Image registration.

Goal: Introducing methods and applications of 2D/3D image registration
- Image based registration: rigid/nonrigid transforms, difference measures, interpolation methods, optimization approaches
- Point based registration: ICP, robust methods, related problems

4. Motion analysis

Goal: Introducing the computer vision techniques used to recover motion from image sequences.
- Motion field and optical flow
- Optical flow algorithms: block matching, Lucas-Kanade

5. Texture analysis

Goal: Introducing texture analysis and methods to extract features and characterize tissues appearance in diagnostic images
-Texture analysis basics
-Texture features: Gray Level Co-Occurrence Matrices. Run Length Matrices, Wavelets
-Supervised classification

Assessment methods and criteria

Written test and project assignment.